A Moral Defense of Smith-RFRA:

Reasonable Pluralism, Political Toleration, and Religious Liberty

Bryan T. McGraw

Associate Professor of Politics & International Relations

Wheaton College

Bryan.mcgraw@wheaton.edu

btmcgraw@gmail.com

mailto:Bryan.mcgraw@wheaton.edumailto:btmcgraw@gmail.com

1

There is a great deal Americans agree about when it comes to religious liberty. The state should not

coerce someone into a profession of faith. Churches, or other religious organizations, should not have the

authority to decide matters of public policy for everyone. And so on. But of late it seems as though our

disagreements have become more widespread and intense, a fact reflected in the continuing disagreements

over how to interpret and apply the First Amendment’s religion clauses. Not only do ordinary citizens

and scholars alike disagree about the outcomes of particular cases. We also disagree about the principles

that should inform how courts, legislatures, and executives understand and decide those cases. More

ominously, to my mind, we also increasingly disagree as to whether we should care much about religious

liberty at all or if we should think of it merely as a species of conscience protections or even as an

obstacle to the construction of a more just society. There is, of course, nothing new or even especially

worrisome about our disagreements regarding the meaning and contours of religious liberty; that merely

makes it of a piece with all the rest of our “fundamental” liberties. But the sense that it may be costlessly

encompassed within a sometimes nebulous right to “conscience” or even dismantled (or severely pruned

back) in favor of other liberties looks to me, at least, to rely on a misunderstanding of what religious

liberty actually is and, more broadly, misconstrue how liberal political orders can best (and always

imperfectly) protect the full range of fundamental liberties under conditions of wide moral pluralism.

This paper engages these questions by constructing an argument how we might best understand and

protect religious liberty as an expression of political toleration and how that understanding reinforces a

moral defense of what I call the Smith-RFRA regime of religious liberty. Federal law and jurisprudence

with respect to religious liberty are generally governed, on the one hand, by the majority opinion in

Employment Division v. Smith and, on the other, by Congress’s legislative response, the Religious

Freedom Restoration Act (RFRA).1 In Smith, the Supreme Court concluded that so long as the state did

not intend to disadvantage any particular religious group (or religious groups in general) with a law or

regulation, it would not violate the First Amendment even if it turned out in fact that some group’s or

individuals’ religious practices were significantly burdened. Concerned that such a rule would do too

little protect religious liberties, Congress overwhelmingly passed RFRA and directed that laws and

regulations had to meet a much higher standard in order to pass muster.2 Federal laws now must not

intend to burden any particular religious practice, and those that do so “substantially” are only legitimate

if they serve a “compelling state interest” and are “narrowly tailored” to serve that interest.

This regime, though broadly popular since RFRA’s passage, has become increasingly controversial

as courts have attempted to apply it to cases involving non-discrimination laws and the Obama 1 Employment Division, Department of Human Resources of Oregon v. Smith, 494 U.S. 872 (1990). 2 RFRA originally covered both federal and state actions, but the Supreme Court struck down the latter in Boerne v. Flores 521 U.S. 507 (1997).

2

administration’s decision to require comprehensive contraceptive insurance coverage under the

Affordable Care Act (ACA).3 While it is eminently understandable that the Smith-RFRA regime would

be controversial, it deserves our moral support, I argue, because it broadly instantiates the most plausible

conception of how we should make our commitments to moral and religious toleration politically

effective under conditions of wide (and reasonable) moral pluralism.4

I develop this claim as follows. I begin with an argument about the nature of toleration itself, settling

on the view that toleration emerges out of the practice of moral judgment in which we adjudicate among

competing and irreconcilable goods with a sharp eye on how those judgments engage others’ individual

agency. I then show that this means that a liberal polity should not only aspire to be generally neutral—

not favoring one way of life over another—but also to try and ameliorate the negative effects of its always

only aspirationally “neutral” institutions, what I (following any number of others) call “substantive

neutrality.” This, I suggest, points us to the moral contours of the Smith-RFRA regime. I then defend this

regime morally against some of its more recent critics who think it imposes unfair burdens on third

parties, concluding that pursuing “clean” solutions to the dilemmas of religious liberty misconstrues how

liberal orders can and should engage the sorts of religious and moral pluralism extant today. The upshot

is that properly protecting religious liberty, like all efforts at political toleration, is a difficult and

politically controversial issue, one whose always partial resolution relies on fallible, limited moral

judgments involving important goods in serious and persistent conflict. Taking refuge in the fantastical

notion that we can resolve these conflicts without remainder if we can just get our principles right is not

just a philosophical mistake; it is, more importantly, a foolish impulse that makes our democratic order

less just and, I worry, less stable.

Toleration and Moral Judgment

We first need to get a clear fix on what toleration means, for its meaning appears both intuitively

obvious and also widely disputed. Generally speaking, when we talk about toleration we tend to have in

mind the idea of putting up with some practice or belief that we find morally objectionable or

inappropriate.5 This suggests why toleration often gets tied to individual autonomy, since to object

3 I have, of course, in mind cases involving objections to the extension of anti-discrimination laws to include sexual orientation and gender identity as well as the numerous (and continuing as I write) religious objections to contraceptives, exemplified by Burwell v. Hobby Lobby, 573 U.S. __ (2014). 4 Just to be clear, I am concerned with the moral contours of Smith-RFRA, not with the legal and constitutional questions per se. It might be the case that Smith-RFRA is morally superior to the alternatives but still constitutionally suspect. 5 David Heyd, "Introduction," in Toleration: An Elusive Virtue, ed. David Heyd (Princeton, NJ: Princeton University Press, 1996). Of course, we also say that we tolerate things that don’t quite rise to the level of being morally objectionable. I can be said to ‘tolerate’ a friend’s perpetual lateness and the objection can be more one of

3

morally to others’ actions or beliefs is to focus inevitably on their capacity to act on their own accord, on

their individual agency and autonomy.6 For some, autonomy is entirely the point. Hans Oberdiek argues

that toleration is necessary so that men and women can pursue their “self-directed li[ves],”7 updating John

Stuart Mill’s justly famous argument in On Liberty, where he looks to maximize political and social

toleration as a means of protecting and encouraging human individuality. One way of characterizing

toleration, then, is as a way of simply protecting individual autonomy.

Individuality is an important good (especially in the context of our society) but making toleration the

handmaiden of autonomy ultimately relies on the unpersuasive claim that a comprehensive, maximal

autonomy limns the edges of what should count as a flourishing human life. As Susan Mendus (among

others) has pointed out, this might require some rather unsavory and illiberal practices to make those who

are incapable of exercising their individuality or for some reason refuse to do so become capable or

willing.8 Overemphasizing autonomy could also so narrow toleration’s scope that it undermines its

practical defensibility. In the course of criticizing Will Kymlicka’s autonomy-based claim for group

rights, Moshe Halbertal argues that focusing on what is “important and central in…life” better describes

the harm that befalls the objects of (unjustified) intolerance than a focus on what it is putatively “chosen”

does. On his account intolerance harms us by “robbing [us] of the possibility of continuing a way of life

that harbors great meaning for [us] as individuals”9 rather than just undermining our ability to choose

some way of life. If the toleration-for-autonomy argument is correct, the many aspects of people’s lives

that are often unchosen – family, religion, language, etc. – might not actually generate toleration claims.

We should recognize, though, that this does not mean that autonomy is unimportant. Halbertal cites

the story of Paul’s conversion to Christianity on the road to Damascus (as related in the book of Acts,

chapter 9) and points out that Paul’s conversion was not simply the product of rational reflection—to

“tolerate” such a conversion simply in terms of autonomy misdescribes what is being tolerated. But that

annoyance than the idea that habitual lateness represents some kind of moral failure. We can, I think, ignore for now this sort of tolerance, since its political impact is rather negligible. 6 This is why I think it’s a mistake to talk about toleration in connection with things like race and sex. These are not chosen characteristics and they are not things we choose. We might talk about tolerating behaviors or tendencies associated with race or sex, but then we are talking about behaviors or tendencies that can, in some sense, reflect our capacity as agents. 7 Hans Oberdiek, Tolerance: Between Forbearance and Acceptance (Lanham, MD: Rowman & Littlefield Publishers, 2001), vi, 120.. See also Emily R. Gill, Becoming Free: Autonomy and Diversity in the Liberal Polity (Lawrence: University Press of Kansas, 2001), Will Kymlicka, Multicultural Citizenship: A Liberal Theory of Minority Rights, Oxford Political Theory (New York: Clarendon Press, 1995). 8 Susan Mendus, Toleration and the Limits of Liberalism, Issues in Political Theory (Atlantic Highlands, NJ: Humanities Press International, 1989), 65-66. See also Joseph Hamburger, John Stuart Mill on Liberty and Control (Princeton: Princeton University Press, 1999). 9 Moshe Halbertal, "Autonomy, Toleration, and Group Rights," in Toleration : An Elusive Virtue, ed. David Heyd (Princeton, NJ: Princeton University Press, 1996), 109-10. This argument is similar to Chandran Kukathas, The Liberal Archipelago: A Theory of Diversity and Freedom, Oxford Political Theory (Oxford: Oxford University Press, 2003).

4

does not mean that Paul had no choice in the matter, a point that Halbertal seems to miss. After his vision

on the Damascus road, after all, Paul traveled on to Damascus, joined the Christians there, and eventually

began preaching, putting his own life in danger. At each point along the way, he had and made a choice

(and if tradition is to be believed, paid for those choices with a martyr’s death). We “tolerate”

conversions because they attach to something in us that is amenable to change, even if “choice” does not

exhaust what happens with conversion. Thus, we should recognize that toleration is indeed inextricably

tied to moral evaluation and individual autonomy, even if the latter does not exhaust what is at stake in

toleration.

Tolerance is unnecessary in two sorts of situations: (a) where everyone agrees on what is moral and

immoral, and (b) if everyone agrees that nothing is moral or immoral. If (b) is true, there is no need to

even talk about tolerance, for there is nothing that can be the object of moral opprobrium.10 But that is

also the case if (a) is true, for if everyone actually agrees on what is moral, then no one can rightly object

to being coerced on those moral grounds. Some seem to anticipate just that situation, suggesting that

toleration is an inherently paradoxical and incoherent concept, less a virtue and more a relic of outmoded

moral judgmentalism.11 Bernard Williams, for one, has famously noted,

The difficulty with toleration is that it seems to be at once necessary and impossible. It is

necessary where different groups have conflicting beliefs - moral, political or religious - and

realise that there is no alternative to their living together. [It is impossible] because people find

others' beliefs or ways of life deeply unacceptable. In matters of religion, for instance..., the need

for toleration arises because one of the groups, at least, thinks that the other is blasphemously,

disastrously, obscenely wrong...It is because the disagreement goes this deep that the parties to it

think that they cannot accept the existence of the other. We need to tolerate other people and

their ways of life only in situations that make it very difficult to do so. Toleration, we may say, is

required only for the intolerable. That is its basic problem.12

Toleration thus seems, even at its best, a puzzling conjunction of claims. We are asked to tolerate what is

“intolerable,” to allow that which we find immoral and wrong to persist or even thrive. If something is

truly wrong, then how can we be asked to tolerate it? We might accept toleration for now on prudential

grounds but it is always only temporary until we come to a resolution on what is actually the case

morally.

The claim that toleration is paradoxical rests on the conjunction that we are asked to tolerate the

‘intolerable,’ and put that way, it does seem like a paradox. But the conjunction looks that way only

10 Bernard Williams, "Tolerating the Intolerable," in The Politics of Toleration: Tolerance and Intolerance in Modern Life, ed. Susan Mendus (Edinburgh: Edinburgh University Press, 1999). 11 David Heyd, "Is Toleration a Political Virtue?" (2004). 12 Williams, "Tolerating the Intolerable," 65. See also Forst, “Toleration and Democracy.”

5

because it elides the distinction between what is ‘wrong’ and what is ‘intolerable.’13 The two are not the

same thing. Or, at least, they need not be the same thing. We can make sense of toleration by realizing

that it is precisely about understanding when, if ever, the two ought to be separated. A decision to tolerate

(or not) does not emerge directly from the judgment as to whether some practice is right or wrong.

Rather, toleration emerges from it being right or wrong in a certain way or in a certain context. Consider

the wrong of adultery. Our fairly libertarian culture still regards adultery as a moral wrong.14 But we

tolerate it legally (and to a lesser degree culturally) except in the military, where service-members can be

prosecuted if a commanding officer judges that adulterous relations have harmed unit effectiveness. Our

willingness to let adultery slide in everyday legal proceedings is not premised on adultery’s moral

acceptance; neither is its legal prohibition in the military a sign that they are more morally attuned than

the rest of us to adultery’s wrongness. Rather, the distinction reflects a contextual judgment about the

circumstances in which a particular wrong ought to be deemed intolerable. If toleration is indeed

conceptually incoherent or paradoxical, then it has to be shown – not assumed – that our everyday

distinctions between ‘intolerable’ and ‘wrong’ do not in fact make sense. To practice toleration is, in

effect, to deny Rousseau’s dictum and attempt to live with one another precisely even as we think each

other condemned (or simply deeply wrong morally).

Preston King seems right, then, when he suggests that in practicing toleration what we are doing is,

in fact, balancing a set of “competing objections.”15 We may, for example, object to someone’s religious

beliefs, either because we think them wrong or morally noxious, but we tolerate them – when we do –

because we think that the objection to denying someone her religious liberty is stronger. If those beliefs

include something that we find entirely unacceptable – say, sacrificing infants – then the religious liberty

claim loses out and we think ourselves justified in practicing intolerance.16 The “difficulty” with

tolerance lies not in some set of internal contradictions or incoherence but more simply in the kind of

judgments necessarily implicated in the very possibility of toleration (and its denial).

Consider that any activity is oriented toward some good (or end), and that some of those goods

conflict; in choosing to pursue one good, we may foreclose the possibility of achieving another.17 When

reflecting on whether to tolerate some practice or belief, central among the competing claims we must

inevitably consider is our own commitment to respecting another person’s agency and autonomy. Again,

13 I should note here that this elision is not a part of Williams’ argument. 14 [Pew research data] 15 Preston T. King, Toleration (London: Frank Cass, 1998). 16 As William Galston has remarked in a different context, there is no religious liberty for the (human sacrificing) Aztec. See his Liberal Pluralism. 17 Isaiah Berlin has, of course, written extensively on this. Ironically enough, so have a number of very much non-liberal natural law theorists. See John Finnis, Natural Law and Natural Rights (New York: Oxford University Press, 1980), Robert P. George, Natural Law, Liberalism, and Morality: Contemporary Essays (New York: Oxford University Press, 1996).

6

in even contemplating the possibility of toleration, our attention is inevitably drawn to the fact that those

participating in that purportedly noxious practice have a kind of freedom to order their affairs as they see

fit, within limits that are themselves, of course, framed by the boundaries of toleration. We need not be

full-blooded Kantians or Millians in our views of human autonomy but in even considering the possibility

of toleration we are already implicitly embracing (or at least taking seriously) a central feature of modern

liberalism—namely, that in deciding whether to tolerate or not, the decision inherently involves taking

serious consideration of others’ capacity to make their own decisions and be responsible for them.18

Toleration is best understood, then, as the emerging from the practice of adjudicating among

competing and apparently incompatible goods in the context of behaviors or views that we find morally

objectionable where we keep a sharp eye on how our judgment takes account of others’ individual

agency.19 But note something important about where the argument has led. It is not toleration itself that,

strictly speaking, interests us. We are interested in, rather, the practice of moral judgment, in the

adjudication of “competing objections” or competing goods, as I have put things. Heyd is right in some

sense, then, to say that toleration is not really a virtue.20 Unlike courage, deciding to tolerate someone or

something is not always appropriate. Similarly, when Oberdiek complains that toleration is inherently

unstable, the instability does not come from toleration per se but from the difficulty in getting our moral

judgments right.21 The fact that so few, if any, have explicitly grasped this fact about toleration perhaps

explains why so many seem so perplexed about it. Getting toleration right means getting our practices of

moral judgment right.

That is, I hardly need say, a condition that makes getting toleration right very difficult indeed. Not

only is it intrinsically difficult to collect and evaluate all the information relevant to any particular case,

but it is also clear that we should not expect that we can articulate reasonably straightforward, clear

arguments that will lead to a broad consensus on contentious moral issues.22 These difficulties might, as

18 Chris Eberle calls this “recognition respect” in Christopher J. Eberle, "Religion and Liberal Democracy," in The Blackwell Guide to Social and Political Philosophy, ed. R L Simon (New York: Blackwell, 2002). 19 At least in this way, this view lines up with the objections Waldron levies against typical liberal toleration claims in Jeremy Waldron, "Toleration and Reasonableness," in The Culture of Toleration in Diverse Societies: Reasonable Toleration, ed. Catriona McKinnon and Dario Castiglione (New York: Manchester University Press, 2003). I say that the goods we consider are “apparently” to be incompatible in the sense that we think of them in the moment as incompatible, even if at some point they turn out not to be. 20 David Heyd, "Is Toleration a Political Virtue?," in Toleration and Its Limits, ed. Melissa Williams (New York: New York University Press, 2008). Or, at least, it is only a second-order virtue. 21 Oberdiek, Tolerance. Indirectly, I think this is part of what Rainier Forst is getting at in describing toleration as “normatively dependent.” We can only understand what it is we are to tolerate (or not) on occasion of some other, more basic, normative principle. For Forst, that is tied to a claim of “unconditioned respect” which cashes out, more or less, Rawlsian terms. See Forst, “The Limits of Toleration.” 22 Think here of Rawls’ “burdens of judgment,” where he quite rightly suggests that our exercise of reason under conditions of freedom will not (nor should we expect) eventuate in even rough consensus for reasons that do not impugn individuals’ motives (e.g. meaning that our disagreements do not merely come out of “bad faith” or the like).

7

some would have it, be partially solved by erecting and enforcing deliberative rules ala Rawls’ “public

reason.” But such rules, as Rawls himself discovered, are notoriously difficult to police and often end up

tying their proponents into argumentative knots.23 Steven Smith has nicely shown how even those

committed in principle to these kinds of rules all too often end up “smuggling” in controversial moral

claims that violate the rules in practice.24 Public reason (or similar kinds of schemes) cannot rescue us

from the difficulties attendant to public moral judgments and get us to the sort of consensus that would

make toleration unproblematic.

So what to do? I hardly have the space or probably even the expertise to offer anything but the barest

sketch of what the practice of moral judgment consists in. But consider how we negotiate these sorts of

issues in our common conversations. As best I can tell, ordinary persons do often make principled, even

philosophical, arguments in the context of moral disagreement, even if those arguments are not as

coherent or tightly constructed as we would sometimes like. But these often tend to run out (in the sense

that they do not definitively settle the issue) and sometimes do not seem to wield much in the way of

persuasive power. We deliberate and we would like to persuade, but I think it is not uncommon to find

that we simply fail to do so. So we tend to make recourse to a different set of resources: stories and

examples (hypothetical and otherwise), and not just in our common conversations. Even in the context of

the most rigorous and abstract moral argumentation, we almost always turn to stories and examples as a

means of persuading others. Charles Larmore points out that even Kant did this,25 suggesting that even at

our most philosophical we are, for better or worse, just simply the sorts of creatures whose moral

judgments are tied in some important way to stories and examples.26 Moral reasoning is not just stories

and examples, but we cannot do without them. This is especially true when it comes to our public

political deliberations.

23 See March, “Rethinking Religious Reasons in Public Justification.” March’s argument—which essentially tries to suggest that there are some sorts of religious arguments that are ok and others that are not—has some real merit, but I suspect is impossible to police well. See also Jurgen Habermas, An Awareness of What Is Missing: Faith and Reason in a Post-Secular Age, trans. Ciaran Cronin (Malden, MA: Polity, 2010). Here, Habermas expands his earlier strict rules on religious reasons in public life, mostly, it seems, because patterns of secularization which he took to be inevitable turned out not to be so. But out of a concern for pluralist democracies’ legitimacy, which he takes to depend in part on laws being justified by secular arguments of a certain type, he suggests authorizing parliamentary leaders to expunge from the official record legislators’ arguments that have religious components. I leave it to the reader to reflect on the strangeness of such a proposal. 24 Steven D. Smith, The Disenchantment of Secular Discourse (Cambridge, Mass.: Harvard University Press, 2010). My own contribution to the debate can be found in chapters 2-4 of my Faith in Politics (New York: Cambridge University Press, 2010). 25 Charles E. Larmore, Patterns of Moral Complexity (New York: Cambridge University Press, 1987). 26 This is, I take it, part of what we are to learn from those who emphasize “narrative” and “tradition.” See Stanley Hauerwas, A Community of Character : Toward a Constructive Christian Social Ethic (Notre Dame, Ind. :: University of Notre Dame Press,, c1981), Alasdair C. MacIntyre, After Virtue : A Study in Moral Theory, 3rd ed. (Notre Dame, IN: University of Notre Dame Press, 2007), Alasdair C. MacIntyre, Dependent Rational Animals: Why Human Beings Need the Virtues (Chicago, Ill.: Open Court, 1999), Alasdair C. MacIntyre, Whose Justice? Which Rationality? (Notre Dame, IN: University of Notre Dame Press, 1988).

8

Why we might be so inclined to argue in this way? Partly, we are interested in how abstract principles

might work out in practice. What makes sense in the abstract can show itself absurd or unpalatable in the

particular. But another part of the answer, and this is the part that Alasdair MacIntyre and Charles Taylor

have been so influential in developing, is also that the moral claims we invoke and the ways in which we

adjudicate among those moral claims (or even decide what counts as a moral claim) are themselves

something embedded in particular communities with particular histories. We need not avert to a crude

historicism to recognize that our moral judgments rely in important ways on the histories within which we

find ourselves. Nor need we avert to a crude Whig view of those histories to recognize how they can

serve to improve our moral judgments. We expect, as a general rule, that individuals with more

experience, more history under their belts, will have better judgment. In dealing with someone who

cannot seem to learn from his mistakes, we shake our heads and, eventually, throw up our hands in

frustration – experience should lead to better judgment. My suggestion is that we tell stories and use

examples in making moral judgments not just as a way of seeing how our principles might apply

practically but just as importantly because they offer us a sort of “moral shortcut” to the benefits history

and experience can otherwise offer. What I mean here is that insofar as hypotheticals and stories help

flesh out our moral reasoning they can function in much the same way that experience can in improving

our judgments.27

Suppose this is right and that our use of stories and examples reflects an intuitive sense that history

matters, both as a relevant set of facts and as an inevitable (and potentially salutary) condition of our

reasoning. If toleration does, as I have argued, reflect the practice of judgment about competing and

incompatible moral goods, then it too is thus tied inevitably to stories and examples – and to history. And

so we come to a picture of toleration that cannot be reduced to some easy or straightforward formula. We

see why it is so difficult, impossible really, to draw a bright line between, say, “self-regarding” and

“other-regarding” actions or between acts that cause “harm” and those that do not. We see why simply

invoking “autonomy” or “liberty” or some-such value cannot solve the problem of toleration, for the

choice to tolerate (or not) depends profoundly on contexts that incline us to value one good over another.

Finally, we see why it is that we so often get our judgments wrong, deciding to tolerate things that we

should not and not tolerating things that we should. Toleration’s critics are right, then, to be suspicious,

but we should be suspicious for the very reasons we should tolerate: we are fallible, limited creatures who

inevitably engage in moral judgments about competing and likely irreconcilable goods, judgments that are

embedded in historical moments and thoroughly suffused with examples and stories meant to enrich those

judgments without any sure expectation that we will always get them right.

27 It probably goes without saying, but we can, of course, go wrong in all sorts of ways precisely by telling and listening to the wrong stories or employing misshapen examples.

9

Political Toleration as Substantive Neutrality

What does this imply for political toleration? First, we should recognize that when we ask this, what

we are asking is how do we structure political institutions and practices so that they make the sorts of

moral judgments described above well, and reliably so? It is not just a question of “scaling up” from our

individual practices, since whatever judgments get made by states come with the promise (or threat) of

coercive force, get written into law, and are typically “stickier” than individual judgments.28 So we

should take correlatively more care in thinking about how to make these judgments: getting our individual

moral judgments wrong might have some worrisome consequences but getting our political judgments

wrong is likely to be comparatively much worse.29

Free societies like ours give a great deal of deference to individual freedom. I suggested earlier that

making individual autonomy a trump for deciding matters of toleration is a mistake, since to do so

erroneously relies on the idea that the fully autonomous life is the singular measure of human excellence.

If it is reasonable to think that the range of good human lives includes those that are not fully

autonomous, and I am convinced that it is reasonable, then human agency and autonomy stand as an

important, but not dispositive good. But it does stand as an important good and we are rightly reluctant to

coerce. This reluctance and the demand for good moral justifications for coercion stand at the heart of the

liberal order, and help show why liberal theorists have long argued that political toleration requires what

they typically call neutrality, the idea that the liberal state should not favor one way of life over another,

insofar as such favoring is not required by the demands of the liberal polity itself.30 To take seriously the

idea that individuals (and groups, for that matter) will work out what counts as a good life differently and

do so reasonably places a significant moral constraint on state actions that promote one conception of the

good life over another.

28 An example here would be the recent Arizona law S.B. 1061, which would have extended protections for religious business owners who wished to refuse certain sorts of services against their conscience, most notably, of course, in the context of same-sex weddings. Some of the popularly voiced opposition was misinformed, but the bill’s opponents understood rightly that if it were passed, it would be much more difficult to undo legislatively. The same is likely true of RFRA. 29 Deciding that your neighbor’s religious beliefs are intolerable will certainly put a damper on next year’s block party, but if the relevant political authority does so, what follows historically is typically much worse. 30 For a rather exhaustive critique of the very idea of neutrality, see George Sher, Beyond Neutrality : Perfectionism and Politics (Cambridge ; New York: Cambridge University Press, 1997). For a response that acknowledges the force of Sher’s critiques but nonetheless shows why neutrality continues to have so much currency, see Andrew Koppelman, "The Fluidity of Neutrality," Review of Politics 66, no. 4 (2004). More broadly, see Koppelman, Defending American Religious Neutrality. Neutrality is, of course, little more than an outworking of the ideas Locke offers up in his Letter Concerning Toleration (1689) where he suggests that the magistrate has no grounds for preferring one sect over another so long as they are not threats to the public order.

10

There are (broadly speaking) two sorts of liberal neutrality, the “neutrality of effect” and “neutrality

of intent.”31 The former argues that state action can be considered properly neutral provided that it has an

equal impact on those affected. “Disparate impact” arguments in legal scholarship more or less embody

this sort of neutrality. Rawls argues that a neutrality of effect is “impracticable” as a matter of

“commonsense political sociology,” noting that it is very difficult – he says it is “futile” – to figure out

exactly how basic political institutions affect the long-term fortunes of distinct ways of life and that in any

case every sort of society involves “social loss.”32 Instead, he offers that neutrality is violated only when

“the well-ordered society of political liberalism fails to establish, in ways that existing circumstances

allow—circumstances that include the fact of reasonable pluralism—a just basic structure within which

permissible forms of life have a fair opportunity to maintain themselves and to gain adherents over

generations.”33 The state is properly neutral, and thus gives practical effect to political toleration, when it

does not as a matter of course intend to disadvantage any particular way of life, at least among those

deemed permissible (or “reasonable”) within liberal democratic societies. Liberal political toleration in

this sense requires that the liberal state avoid importuning those ways of life compatible with a just and

stable liberal democratic political order.

The very fact that Rawls distinguishes his version of neutrality from a neutrality of effect tells us that

some ways of life will inevitably do better than others. He suggests that ways of life permissible under a

liberal polity may still “fail to gain adherents” and perhaps even disappear, but that such an outcome,

while it might be lamentable, does not necessarily rise to the level of injustice. As long as a liberal

society’s basic institutions are not “arbitrarily biased” against those ways of life and that the conditions

leading to their decline or demise are not themselves unjust, there is little more to be said.34 Others have

not been quite as sanguine, suggesting that this sort of benign neglect often amount to an injustice.

Instead, they argue that a liberal state has the responsibility to ensure that groups of citizens, especially

those who have historically been the object of injustice, feel properly included in the wider political

community. Rather than neutrality, or toleration more broadly, the state should pursue what they call

recognition.

Recall that one of the things essential for the practice of toleration was critical moral evaluation. In

order to say that you tolerate something, you need to think it morally objectionable. So to the degree that

liberal neutrality is meant to instantiate a regime of political toleration, it suggests (even if only 31 Koppelman helpfully catalogues a wider range of the ways in which neutrality gets used in liberal political thought, but for our purposes here, it is enough to focus on intent and effect. 32 John Rawls, Political Liberalism, 2nd ed. (New York: Columbia University Press, 1996), 193-200. 33 Rawls, Political Liberalism, 198. I should note that Rawls expresses some discontent with the term “neutrality” because it inevitably seems to call to mind the idea of neutrality of effect. See John Rawls, Collected Papers, ed. Samuel Richard Freeman (Cambridge, MA: Harvard University Press, 1999), 460. Nonetheless, he continued to use it. 34 See Rawls, Political Liberalism, 196-99.

11

implicitly) to those whom it “tolerates” that their acts or views are thought by the state (or perhaps just by

most of the citizens at large) to be problematic in some respect: the liberal state tolerates what society

disfavors. Since those being “tolerated” in a liberal democratic society are likely also to be largely

excluded from significant social or political influence – otherwise, they would hardly need political

toleration – they are likely to find themselves doubly marginalized. This, as Anna Galeotti points out, has

the practical effect of undermining their civic standing. Being the object of political condescension

makes it difficult to develop the kind of self-respect necessary for the effective exercise of civic duties

and even for pursuing human flourishing.35 Rather than simply encouraging mutual toleration, the state

should, she says, “make all citizens feel positively at ease with their full-blown identities in public as well

as private.”36 It should extend, through subsidies, legal protection, or some sort of affirmative cultural

policy, political recognition: state action designed to provide material and psychological support for the

disparate and competing religious, ethnic, or cultural identities that inhabit modern democratic life.

Recognition has the real virtue of taking account of the ways in which an ostensibly neutral set of

political institutions can serve to reinforce or even create social inequalities that can then have pernicious

political consequences. But it has real problems as well. It oversimplifies the connection between civic

capacities and cultural esteem37 and neglects the ways in which mediating institutions like families,

churches, and parties can themselves create and sustain cultural identity, making state action less

necessary and more problematic.38 It can very quickly, as both Taylor and Galeotti recognize, run afoul

of even the most basic liberal commitments.39 Arguments in favor of recognition also seem to ignore or

(at best elide) the zero-sum quality of some political conflicts: extending recognition to one group

sometimes means de-recognizing another.40

35 Anna E. Galeotti, Toleration as Recognition (New York: Cambridge University Press, 2002), 12. 36 Galeotti, Toleration as Recognition, 105. 37 For an argument that American evangelicals thrive in part on the perception that they are an “out-group,” under siege from the broader culture see Christian Smith and Michael Emerson, American Evangelicalism: Embattled and Thriving (Chicago, Ill.: University of Chicago Press, 1998). Such perceptions reinforce subcultural boundaries and can serve to police inter-cultural disputes. 38 Nancy L. Rosenblum, Membership and Morals: The Personal Uses of Pluralism in America (Princeton: Princeton University Press, 1998), 329-31. I say more problematic because political attempts to subsidize cultural identity necessarily has the effect of selecting one version of that identity and probably has the effect of subtly changing that cultural identity in ways that help it secure those subsidies. 39 Taylor declines to endorse the most robust version of recognition, but rather suggests that we should approach diverse cultures with the “presumption” that they are equally worthy of respect and only make judgments to the contrary after we’ve delved more deeply therein. For her part, Galeotti suggests that recognition could be “symbolic” and thus avoid running afoul of basic liberal commitments. 40 Consider something as relatively innocuous as public holidays. In deciding who or what gets honored with holidays, the state goes some way toward ‘recognizing’ one element of a pluralistic culture over another. To the degree that the state decides to adjust those holidays to shift its recognition, it is likely that some groups will feel slighted as they ‘lose’ their holiday.

12

Perhaps most importantly, the argument for recognition founders on a tension that stands at the heart

of its claims. It claims that in modern democratic societies there exist diverse and contesting ways of life

on whose fortunes purportedly neutral laws will have disparate – and unjust – effects. But it then argues

that to compensate for (or to avoid) such injustices, laws should be altered and public resources

committed to ameliorating those effects. What is lacking is any reasonable argument as to why or how

those laws will be altered or resources committed. Why will the members of these diverse and contesting

cultures agree to extend to one another these sorts of resources if, in fact, they are as diverse and

contesting as the recognition argument suggests? If the marginalized cultures are really as marginalized

and despised as advertised, it is hard to see how a practical program of recognition gets off the ground at

all, except to the degree that these diverse and contesting ways of life are actually not all that diverse or

contested. If, as Galeotti put it, the goal of recognition is to “make all citizens feel positively at ease with

their full-blown identities in public as well as private,” that can only happen so long as those identities are

relatively easily compatible with one another. Recognition tends to involve the extension of public

benefits (money, employment set-asides, legal exemptions, and so on) to minority groups that have

historically been the object of social marginalization and ostracism. These benefits must garner, I take it,

some reasonably significant levels of political support. It must be the case that some large (or at least

politically potent) portion of the population clearly accepts the necessity of extending recognition and

most citizens must already view minority groups at least somewhat positively, meaning in effect that

problem for which recognition is meant to be a solution is pretty marginal itself. After all, if some group

is so widely despised that they would qualify for recognition, it is hard to see how they would get it, and

if they get it, do they really need it?41

In any case, recognition looks to me to be unpersuasive as a full-blown alternative to neutrality or

political toleration. Thinking through the recognition claim does, however, point up something important.

The arguments for recognition are rooted in the commonplace observation that institutions can

systematically and predictably disadvantage groups of people even if they are not meant to do so. The

claim for recognition takes that to mean that the whole edifice of neutrality ought to be overthrown.

Rawls (among others) argues, alternatively, that the neutrality of intent is the right way to make toleration

politically effective because it fulfills what justice requires and is, unlike the neutrality of effect and

recognition, practical. The neutrality of intent fulfills the requirements of justice because it treats various

ways of life fairly. The fact that some ways of life prosper and others fall into desuetude is not, he thinks,

41 An exception to this would be recognition policies put in place by courts or bureaucracies largely insulated from popular sentiment. But as the fairly widespread critiques of multiculturalism in Britain shows (critiques voiced on both the left and right), if groups really are unpopular, it seems reasonable to think that democratic majorities will usually eventually get their way. Not always, certainly, but usually.

13

in itself a problem with a neutrality of intent. As long as the institutions do not aim at promoting one way

of life over another, no one is done an injustice, even if her way of life does not prosper.

Suppose, though, that some people find it systematically more difficult than others to pursue their

way of life because liberalism’s political institutions exact what John Tomasi has called an “unequal

psychological tax.”42 That is, even though their political views qualify them as citizens in good standing,

these citizens find themselves persistently frustrated in pursuing their non-political ends. As Rawls points

out, ways of life might do poorly under political liberalism for one of two reasons. First, they might be

“in direct conflict with the principles of justice.”43 Slaveholders will find their lives frustrated in liberal

societies, but, of course, that is nothing to be concerned about; the fact that ways of life dependent on

injustices incompatible with any reasonable sort of liberal politics will not do well under liberal

institutions is part of reason for upholding those institutions in the first place. Second, some ways of life

“may be admissible but fail to gain adherents under the political and social conditions of a just

constitutional regime.”44 Examples of this type include “certain types of religion” that “can survive only

if [they control] the machinery of state and [are] able to practice effective intolerance.”45 Since this sort of

religion could only survive provided that the state violates a central liberal claim – religious liberty – its

passing (or radical change) might be lamented, but cannot be considered unjust.

Rawls’ second example here is a strange one, for it does not differ in principle from the first. In both

cases, a way of life or comprehensive view can only be sustained through a political injustice that any

liberal democratic regime is fundamentally committed to preventing.46 More to the point, what does

Rawls mean by saying that the coercive religion is “admissible” under a liberal regime? Perhaps he

means that a liberal state need not outlaw such a religion, as opposed to its duty to outlaw slaveholding,

but in the context of trying to understand whether political liberalism is fair to different ways of life, it is

an unhelpful example, to say the least. The question that needs addressing revolves around when political

institutions systematically disadvantage some people over others and whether they do so for good

reasons. So if the establishmentarian religion does poorly, then by definition there is not a problem, but

that does not mean much more than saying that political institutions animated by liberal principles are

42 John Tomasi, Liberalism Beyond Justice (Princeton: Princeton University Press, 2001), 35. He presses roughly the same logic in his recent Free Market Fairness (Princeton, 2012), except there he is more interested in the relation between property rights and flourishing than cultural matters per se. 43 Rawls, Political Liberalism, 196. 44 Rawls, Political Liberalism. 45 Rawls, Political Liberalism. 46 As Alfred Stepan has pointed out, of course, an established church is not in itself incompatible with liberal democratic government. See Alfred Stepan, "Religion, Democracy, and the 'Twin Tolerations'," Journal of Democracy 11, no. 4 (2000). A coercive establishment, where membership is compulsory, is another matter altogether.

14

unfavorable to those who oppose them. It says nothing about believers whose beliefs and practices are

broadly compatible with liberal politics.

This lands us at what I think is a quite important question: should there be anything be done if

liberalism systematically disadvantages ways of life that are compatible with liberalism properly

understood? Is the neutrality of intent truly fair if it has predictably and persistently disparate effects?

Liberalism seems committed to being fair to the different ways of life within its borders and if it looks

like some of those ways of life do poorly because of liberalism’s institutions, then it seems quite plausible

to say that we should at least explore ways to make liberalism’s requirements less burdensome or more

equitable.

Rawls declaims any such attempt. He rightly argues that any set of social institutions will

necessarily favor some ways of life and disfavor others. No one could reasonably expect to achieve a full

“neutrality of effect,” even if that were desirable. But Rawls goes further, and suggests that it “is futile to

try to counteract these effects and influences, or even to ascertain for political purposes how deep and

pervasive they are. We must accept the facts of commonsense political sociology.”47 This is a strange

conjunction of claims. We must, Rawls tells us, “accept the facts of commonsense political sociology,”

which would apparently have us (a) believe (and rely on the idea) that the effects of political institutions

are deep and pervasive and (b) accept that it is futile to try and counteract such effects when we deem

their unintended consequences unfortunate in some respect. If we take Rawls at his word here, then we

are left with the rather puzzling idea that we know that political institutions have lots of non-political

effects, but we cannot really know what those effects are – or, at least, how “deep and pervasive” they are

– and we certainly cannot do much of anything about them, except, it seems, rely on them to underpin our

liberal political institutions.48

It is almost certainly true that we cannot discern precisely how political institutions affect social and

private life, and it is also true that our ability to design political institutions to effect particular goals is

always only partial, but we can still say a great deal about how institutions do in fact have their effects. If

all shops are closed on Sundays, Christians probably have a comparative advantage over religions that

have their worship services on other days. If all children are forced to go to secular schools, probably

fewer of them will be religious than if some had gone to the religious schools of their (religious) parents’

47 Rawls, Political Liberalism, 193. 48 Rawls notes, in a discussion of public reason, that his argument for political liberalism relies at least in part on the “hope…that the political conception and its ideal of public reason are mutually sustaining, and in this sense stable.” They can be so as long as “A well-ordered society publicly and effectively regulated by a recognized political conception fashions a climate within which its citizens acquire a sense of justice inclining them to meet their duty of civility and without generating strong interests to the contrary”(1996, 252). In other words, political liberalism works as long as political institutions are as effective as he deems them to be in producing ‘just’ citizens. His argument is thus triply puzzling. A cynic might suppose that Rawls retreats to the citadel of “ideal” theory and declaims any empirical judgments when it is especially helpful to his argument.

15

choice. More to the point, even if no one intended for a particular set of institutions to be unequal in their

effects, once such effects are evident, refusing to adjust those institutions to account for those effects

clearly then becomes part of “intent.” “Commonsense political sociology” can tell us a great deal about

how institutions shape social life, and even if it is far from perfect, a set of institutions whose design turns

out to favor one way of life over another cannot really be called neutral, even if they are not at first

intended to act in that manner. And to the degree that the possible adjustments are themselves compatible

with liberal principles and not productive of greater problems, political institutions absent those

adjustments cannot be considered properly neutral or fair.

So it seems to me that Tomasi is right to suggest that once liberals have committed themselves to a

neutrality of intent as a means of making toleration politically effective, they are then in turn committed

to ensuring that such neutrality is truly as neutral as it can be; they must be concerned with political

institutions’ non-political effects.49 But Rawls and others might complain that this simply leads us back

to some kind of neutrality of effect, a standard that he plausibly claimed was “impracticable.” It is quite

difficult to delineate with sufficient clarity whether some people’s lives go well and others go poorly due

to political institutions, technological shifts, cultural trends, or any of the other bewilderingly myriad

ways that societies change. Moreover, it is not even often clear whether people’s lives are going well,

going poorly, or simply changing. Multiplying these uncertainties across groups of people produces some

real sympathy for Rawls’ (and others’) unwillingness to tread in such territory.

Nonetheless, it is not so obvious what sort of force the “practicability” objection should have,

especially when we are considering disparate effects on ways of life politically compatible with

liberalism. If, for instance, political institutions could be arranged so as to have perfectly equal effects on

the various ways of life compatible with political liberalism, what reason could there be for not doing so?

What is striking about Rawls’ discussion of the neutrality of effect is the absence of a moral argument: the

critique is entirely practical. There is nothing in Rawls’ argument to suggest he thinks there is anything

morally wrong with thinking about neutrality this way, it is just too hard to make it work. If that is right,

then perhaps practicability ought to be understood as little more than a prudential warning against

expecting too much out of a system of political neutrality; it is not a knock-down argument against

attempting to mitigate political institutions’ disparate effects on people’s ways of life. It is perhaps

unavoidable that political institutions will unfairly advantage some ways of life over others, while it is

just as unavoidable that a liberalism properly understood will understand itself as at least having a moral

obligation to assuage that unfairness as a means of making toleration politically effective.

Rather than a neutrality of intent, liberal states seeking to make toleration politically effective could

commit to something analogous to what Stephen Monsma and Douglas Laycock, among a number of 49 Tomasi, Liberalism Beyond Justice. See especially pp. 40-56.

16

others, have called (in the context of discussions of how to interpret the First Amendment) “substantive

neutrality.”50 In this view of neutrality, the liberal state should take care that its actions do not, at first

pass, intend to advantage or disadvantage any particular way of life, insofar as that way of life looks to be

compatible with liberal order. It should then take further action, or reconstruct its first action, to try and

mitigate the disadvantages that might emerge on account of that first action. In other words, the liberal

state should act fairly with respect to most of the diverse ways of life within its borders and look to

correct itself if and when any unfairness becomes evident.

It might be easy to suppose that this substantive neutrality runs into many of the same problems I

suggested above for recognition, and it is true that both can get themselves entangled in any number of

practical problems such that we might be tempted to abandon this amended view of neutrality as well.

Two things distinguish substantive neutrality, though, and suggest why we might endorse one but not the

other. First, recognition requires (or assumes) that citizens possess a kind of positive regard for one

another, while substantive neutrality more plausibly merely asks us to be attentive to the ways in which

state actions put some citizens at a disadvantage and see if there are ways to mitigate that disadvantage.

That is, substantive neutrality does not ask citizens to forego their critical views of others’ lives or beliefs

but instead to construct political institutions such that they are reasonably likely to be properly tolerant of

those whose views do not in the main shape policy outcomes.

Second, note how substantive neutrality, unlike recognition, is an inherently corrective process; it

takes for granted that even well-meaning policymakers will act in ways that substantially favor some

groups over others. One way to think about this is that substantive neutrality calls for liberal states as a

first-order effort to be evenhanded when it comes to distributing their benefits and burdens and then to try

and correct that evenhandedness as a second-order effort when it becomes clear that the first-order

evenhandedness carried with it some inequitable distributions. Recognition hopes for the establishment of

a social order in which are equally valorized by the state and equally (or at least reasonably) valorize one

another. Substantive neutrality has much more modest hopes: we might be able to correct some inequities,

but it is unreasonable to suppose that we can correct them all. Even when political actors act in good faith

this is true because sometimes politics really is a zero-sum game. When we secure one good, we

sometimes must sacrifice another.

This view reflects, as should be clear, how it is I argued above we ought to think about toleration

itself. Toleration emerges from the practice of moral judgment in which we make choices regarding

50 Douglas Laycock, "Formal, Substantive, and Disaggregated Neutrality toward Religion," Depaul Law Review 39 (1990), Michael W. McConnell, "Believers as Equal Citizens," in Obligations of Citizenship and Demands of Faith, ed. Nancy L. Rosenblum (Princeton: Princeton University Press, 2000), Steven Monsma, "Substantive Neutrality as a Basis for Free Exercise-No Establishment Common Ground," Journal of Church and State 42, no. 1 (2000). The context for these arguments is the interpretation of the First Amendment to the U.S. Constitution.

17

apparently irreconcilable goods with an eye on individual agency. In pressing on this second-order effort

of substantive neutrality, we genuinely look to redress inequities precisely because we have a serious

commitment to treating others fairly with regard to their own agency and independence. But cannot

expect to do so without remainder: in securing one set of goods we (nearly) inevitably forfeit others. Such

efforts are quite likely to be messy, imperfect and even on occasion issue in real injustices, but so long as

we inhabit a democratic society marked by reasonable moral and religious pluralism, substantive

neutrality looks to be the best option on offer as a means of making toleration politically effective.

Political Toleration, Substantive Neutrality, and Religious Liberty

My argument to this point has been that we should understand toleration as emerging out of the

practice of moral judgment about competing and apparently irreconcilable goods in which we take careful

notice of others’ agency and independence. That understanding is best put to effect politically in a regime

that attempts to instantiate something like substantive neutrality, where the liberal state commits itself to

both aspire to be neutral in its policymaking and redress deviations from that neutrality so far as its own

commitments permit. In this final section of the paper, I show how this understanding of toleration

supports morally what I call the Smith-RFRA regime of religious liberty and defend it against three

important objections.

The Supreme Court’s post-war religious liberty jurisprudence reflected a kind of studied

ambivalence prior to Employment Division v. Smith. Though the court generally claimed that the First

Amendment required the government to meet a “strict scrutiny” standard if some law or regulation

burdened an individual’s religious belief or practice, it tended to apply that standard somewhat loosely in

practice. In Sherbert v Verner (374 U.S. 398), for example, the court held that South Carolina could not

withhold unemployment benefits because someone refused a job on account of her sabbatarianism. To do

so would be to make individuals violate their religious commitments for ends that were, in the court’s

view, less than compelling. In Lyng v. Northwest Indian Protective Association (485 U.S. 439), the court

decided that the need to build a road was just the sort of compelling interest that could overcome bona

fide religious liberty objections. The upshot is that, prior to Smith, religious liberty protections were, as

the saying went, “strong in theory and weak in fact.” In Smith, Justice Scalia looked to do away with the

ambivalence, arguing for the majority that the First Amendment did not in fact require the state to meet a

compelling interest test because there was no basic constitutional obligation to accommodate religious

practice (though the constitution did permit legislatures or executive agencies to do so on their own). So

long as the state did not intend to disadvantage (or advantage) any particular religious tradition—e.g. so

long as the state acted neutrally with respect to religion—it was more or less in the clear. Concerned that

such a rule did too little to protect those who might not be able to get legislatures to provide those

18

accommodations, Congress overwhelmingly passed RFRA, requiring that federal laws and regulations be

held to the stricter standard.51 This combination is what I call the Smith-RFRA regime of religious

liberty.

To the degree that we think of protecting religious liberty as a particular instance of making

toleration politically effective, it should be clear how the Smith-RFRA regime reflects substantive

neutrality. On the one hand (via Smith), the state is committed to neutrality among religions and between

religion and non-religion, and on the other (via RFRA) the state has acted to redress what it perceived as

likely inequities emerging from a more or less strict application of that neutrality. My point is not to argue

for this regime as the best interpretation of the First Amendment or for its particular applications, but

rather merely argue that the regime’s moral contours reflect well what seems to me to be required to make

toleration politically effective. Not all would agree, of course, and we can flesh out why I find this moral

claim persuasive by considering three lines of criticism, namely that this regime offers unjustifiable

special protections for religion, that it unfairly imposes third-party burdens, and that its reliance on

legislative bodies to correct inequities is morally and politically dangerous. Let me take each in turn.

Though there was broad support for RFRA when it was originally passed, the small number of critics

grew as it became increasingly clear that strictly enforced religious liberty protections might stymie

“progress” with respect to any number of public policy goals, especially those that tended to touch on hot-

button “culture war” issues. Some concluded that the problem lay less with any particular interpretation of

RFRA or like-minded pieces of legislation and more with the notion that the liberal state should be

interested in offering special protections for distinctively religious liberty at all.52 The idea here is fairly

straightforward, namely that in offering heightened protections for religion as such, the liberal state is

acting unfairly with regard to similarly constituted non-religious perspectives and lives, thus making

Smith-RFRA internally incoherent. It is unreasonable to endorse equality on one hand and undermine it

on the other. On this view, RFRA doesn’t correct inequities so much as it unfairly (and unreasonably)

privilege religion. Instead, we should simply protect “conscience” or some-such aspect of human life that

better and more equally covers all, maybe something like what Eisgruber and Sager mean by their claim

in favor of equality.53

In one sense, I think this critique has real merit, at least in the sense that since the border between

religion and non-religion is well nigh impossible to police cleanly, picking out one group of practices and

beliefs for heightened protections looks like treating similarly situated citizens unequally. But in and of

itself, it doesn’t really tell against Smith-RFRA for two reasons. First, insofar as RFRA actually does

51 A good bit of the scholarly disputes about the recent Hobby Lobby decision hang on precisely what Congress meant in requiring the “compelling interest” standard. 52 Cf. Leiter, Why Tolerate Religion?; Schwartzman, “What If Religion Is Not Special?” 53 Eisgruber and Sager, Religious Freedom and the Constitution.

19

work to correct inequities, we should not indict it because it does not correct all inequities. For Leiter and

others, the answer to the equality challenge is to reduce the privileges afforded religious belief and

practice, a kind of leveling down, but why not instead think about ways of protecting other, equally

important commitments that individuals and groups might have? Instead of saying that because it seems

unfair, for example, that religious non-profits (e.g. churches and the like) get special exemptions from

certain kinds of regulatory regimes that non-religious ones don’t, why not just ask whether the similarly

situated non-religious ones should get them as well?54 This leads to a second point, which is that the

reason it is really rather difficult to simply privilege up everyone’s protections is precisely the same

reason that made this a difficult question in the first place, namely the question of how do we identify or

even think about similarly situated non-religious commitment in need of religion-like protections?

Consider the following, somewhat specious, example. Sports fans of all sorts exhibit what we might

think of as religious devotion. They set aside significant amounts of time to watching their teams perform

(either in person or on television), they organize ritual feasts around the sporting events, they spend

significant sums of money to attend those events and dress themselves in the proper vestments, they

venerate holy saints of seasons past, and they engage in lengthy, sometimes tedious, speculations about

their denomination’s prospects and the like. They even engage in theodicy-like speculations if their teams

are perennially bad. For some, their fan commitments clearly occupy a central organizing place in their

lives. Should they get the same sorts of accommodations as traditional religious believers in the form, say,

of accommodations at work? At first blush, the answer seems to be obviously no, but given that we do

count non-theistic religions as religions, it’s not as clear why we shouldn’t. Who’s to say, after all, that

the individual willing to spend large sums of money to sit in below-zero temperatures to cheer on his team

isn’t devoted in such a way that deserves the same sorts of accommodations that standard religions

receive? If we genuinely don’t think we can plausibly pick out religion as something worth protecting as

such, we may very well end up treating Catholics and NFL fans similarly, either by imposing unfair

burdens on Catholics or by validating the exaggerated importance some people attach to spectator sports.

The point here is to merely point out that the equality worry might cut in two different ways. It might

ask us, in the name of treating all equally, to do real damage to religious liberty by leveling things down

or have to defend potentially silly protections or accommodations by leveling things up. Think here of

less specious examples like France’s laïcité (which significantly impinges on religious liberties in an

effort to maintain a secular public sphere) or the erection of Satanist memorials (which are really just

ways to mock mostly Christian memorials on public grounds). The problem is less with the equality

critique as such as it is with the expectation that equality as such can indeed offer enough of a principle to

adjudicate these sorts of judgments well. Indeed, as Andrew Koppelman points out in engaging Eisgruber 54 For a nice argument to this end, see Vallier, Liberal Politics and Public Faith.

20

and Sager’s argument for equality, they only really succeed insofar as they figure out how to smuggle in

the sorts of accommodations of religion that seem reasonable to them—but certainly are not required by

the principle itself.55 No abstract principle, I think, will give us enough to make the sorts of judgments we

must make in these sorts of cases and it is no critique of Smith-RFRA as such that Congress judged that

religious belief and practice needed protections the Constitution (as interpreted in Smith) did not always

provide. Perhaps other views or ways of life deserve protection as well, and if so then perhaps Leiter, et.

al. can push for a NFRA statute, the Non-religious Freedom Restoration Act.

Perhaps that is too easy. For inasmuch as Koppelman, for example, agrees that the effort to make

everything fit under one principle is rather quixotic, he has been, along with a number of others, quite

critical of the Supreme Court’s decision in Hobby Lobby, arguing that by validating Hobby Lobby’s

demands, the Court allowed Hobby Lobby’s owners to impose costs on women who did not share their

convictions.56 We might think of this “third-party” problem as a more specific version of the equality

problem. The basic claim is that the state granting an exemption is legitimate, all things considered, only

so long as in granting the exemption, the state doesn’t impose a “material burden” on third parties. This

doesn’t hold, of course, if the burden is de minimis but in the case of Hobby Lobby, it’s plausible to think

that at least in some cases, women might have to pay significant sums or find employment that did offer

the contraceptive coverage they required (or desired). As with the more general equality claim, this surely

is something to take seriously, but I don’t think it stands as a generalizable critique of Smith-RFRA in the

way its critics think.

Consider that so far as I can tell most of the critics of the Hobby Lobby decision accept and even

endorse the set of exemptions and accommodations the administration has offered to churches and

religious organizations. The basic claim seems to be that because it is more likely that those institutions’

employees agree with the objections, then the third-party burdens are not as much of an issue. That is far

too broad a brush, it seems to me, as it neglects the obvious truth that churches often employ people who

don’t agree with some of their employer’s moral views. If the employees of Hobby Lobby shouldn’t be

burdened why should a church’s maintenance workers?57 Moreover, suppose that the mandate wasn’t for

access to contraceptive care, but for abortions. Does the critique work in the same way? In one sense,

obviously yes—just like with contraception, if an employee’s health plan does not cover abortions, then

she will have to pay for it, thus bearing a financial burden not (strictly) of her choosing. But we tend to

treat abortion differently than contraception. It is much more morally controversial and we are much less 55 Koppelman, Defending American Religious Neutrality, 153–165. 56 See Gedicks and Koppelman, “Invisible Women.” For other articles in a similar vein, see Gedicks, “One Cheer for Hobby Lobby”; Lupu, “Hobby Lobby and the Dubious Enterprise of Religious Exemptions.” 57 Perhaps, of course, you could then make recourse to the for-profit/non-profit distinction, but it’s not at all clear to me that this does much work for you morally, whatever its legal provenance. As tax specialists are wont to quip, the difference between for-profit and non-profit isn’t as stark as we often suppose.

21

likely to coerce people to pay for them, to perform them, etc. And we would be much more likely to offer

blanket exemptions to religious organizations of all sorts (at least in the non-profit world) if abortion were

included in the mandate precisely because we generally judge that the trade-off in that case is worth it.

Perhaps my judgment here is incorrect. But note the form of the deliberations here: we think that some

good is worth securing through political action, but recognize that securing that good has moral costs. We

could simply decide that the moral costs are acceptable and move on or we could (as we often do) try and

find ways of mitigating those costs, keeping in mind that even the mitigating itself has costs that must be

included in our deliberations. We are doing precisely what I suggested we must do with respect to

political toleration, and so even if critics think that any particular burden imposed on others is too great,

they nonetheless are inevitably thinking about the question in ways my view of toleration says we must.

There is no reason, then, to think in general that our moral and political equality is better served by

attempting simply to make recourse to some general and encompassing principle. But this takes us

directly into a third set of concerns, namely that Smith-RFRA implicitly supposes that we are better

served in having our liberties protected by legislative rather than judicial bodies. Because RFRA is an

ordinary piece of legislation that can be modified or abolished as any other piece of legislation, we might

worry that it leaves too much to legislative bodies, which are certainly not particularly reliable, especially

when it comes to issues of religious liberty. RFRA’s proponents were in part motivated by skepticism

regarding Scalia’s assurances in Smith that accommodations could still be offered via legislative or

executive action. It seemed to them (and many of us, no doubt) more likely that those bodies would carve

out exemptions for the politically connected and leave the rest to make the best of it on their own. There

are, I think, legitimate equality worries to be had here and yet I still think that legislative bodies must play

an important role in working through issues regarding the substance (and limits) of religious liberty.

Implicit in my response above to the equality concerns is the idea that in trying to figure out how and

where to draw distinctively corrective lines, we are probably better off in doing that via legislative action.

To put it more in the language of my broader argument, my answer above suggests that when thinking

about whether and how to take corrective action in the context of a broader commitment to political

equality, we may be better off by allowing these corrective questions to be decided by legislative bodies

informed by public political deliberations. It’s easy to see how such a strategy might be worrying to

religious minorities of all sorts; let me explain why it is still the best option on offer.

It strikes me that we can make two opposite mistakes in thinking about this sort of question. On the

one hand, we can ignore this worry and simply trust legislative majorities as such to respect the liberties

of religious minorities. That seems naïve. On the other hand, we might attempt to avoid legislative action

altogether and make courts solely responsible. The trouble with that option is that it likely ends back up in

something like the equality principle. Scalia’s argument in Smith won a majority in part because it offered

22

a justiciable rule that could then be relied upon to make judgments. Courts, especially those at the

appellate level, seem inevitably to incline to the establishment of generalizable rules that can then (for

obvious appellate sorts of reasons) help lower courts decide other particular cases. I have already

suggested why that is a problem, but I think it’s a problem for both strategies. They both rely, if only

implicitly, on the idea that securing religious liberty hangs above all on getting our single principle right

and then ask which institution is best suited to making that principle effective. But if my toleration

argument is correct then that is an implausible view, for it assumes that we can have our political

institutions securely and predictably make the sorts of moral judgments that our condition of reasonable

pluralism necessarily means we cannot.58 What is striking about the Smith-RFRA regime is just how it

does not simply leave the legislative branch free to draw its lines wherever it sees fit, but instead works to

set a certain (compelling interest) standard which then courts must interpret and apply, both in terms of

legislative and jurisprudential history. What’s more, because RFRA is just an ordinary piece of

legislation, it can be overturned or modified via a simple majority, just as Senate Democrats attempted to

do recently with respect to RFRA’s application to for-profit corporations.

This interplay between the branches has a number of qualities to recommend it as a way of

discerning how best to take corrective actions. Perhaps most importantly, it looks to me to reflect just the

sort of historically rooted process of judgment I argued earlier issued in toleration (or its limits). Prior to

Smith, the Supreme Court’s repeated ambivalence regarding how to apply its stated “compelling interest”

reflected the reasonable judgment that actually imposing strict scrutiny would be in practice unworkable.

In passing RFRA, Congress seems to have disagreed, though scholars themselves disagree about how

much Congress in fact merely meant to restore the status quo ante Smith and how much they meant to do

more. In Hobby Lobby, a majority decided for the latter. Perhaps that will indeed turn out to be a mistake

and be unworkable; if so, then Congress can adjust or abolish RFRA, a process that looks to me much

preferable than relying on the insulated intuitions of a court majority. Yes, courts will sometimes get

things wrong, as will legislatures; the institutional interplay of the Smith-RFRA regime looks to me to

provide a good institutional framework for minimizing the likelihood of either and in fact

institutionalizing the correlative interplay between principle and judgment that political toleration

requires.

Conclusion

Religious liberty has become a rather fraught political issue of late, itself perhaps an unsurprising

consequence of what Rawls calls the “fact of reasonable pluralism” combining with the expansion of the

58 A number of scholars have of late made similar sorts of points. Cf. DeGirolami, The Tragedy of Religious Freedom; Smith, The Rise and Decline of American Religious Freedom.

23

regulatory welfare state in post-war democracies. For some, the difficulties attendant to the myriad cases

swirling around our contemporary religious liberty cases as well as the philosophical difficulties in

pinning down exactly what is valuable in protecting religious liberty suggest to some that we would do

well to downplay religious liberty as such and look to protect instead something like “conscience.”

Inasmuch as we think about protecting religious liberty as an aspect of political toleration, though, we

should recognize that making recourse to some allegedly more defensible or simple principle as a way of

making our liberties secure and defensible is to fall for a false promise. Precisely because protecting

religious liberty involves moral judgments among competing and apparently irreconcilable goods—either

you apply non-discrimination norms or you allow an institution to hire according to mission, for

example—we are inevitably already caught up in the sort of process I suggest Smith-RFRA embodies.

These protections for religious liberty are not perfect, but they are the best available and are likely more

stable and secure than the extant alternatives.

24

Bibliography Black, Amy E., Douglas L. Koopman, and David K. Ryden. Of Little Faith : The Politics of George W. Bush's

Faith-Based Initiatives. Washington, D.C.: Georgetown University Press, 2004.

Brighouse, Harry. On Education, Thinking in Action. London ; New York: Routledge, 2006.

Bruce, Steve. Conservative Protestant Politics. New York: Oxford University Press, 1998.

Callan, Eamonn. Creating Citizens: Political Education and Liberal Democracy, Oxford Political Theory. New

York: Clarendon Press, 1997.

Dworkin, Ronald. Freedom's Law : The Moral Reading of the American Constitution. Cambridge, Mass.: Harvard

University Press, 1996.

Eberle, Christopher J. "Religion and Liberal Democracy." In The Blackwell Guide to Social and Political

Philosophy, edited by R L Simon. New York: Blackwell, 2002.

Finnis, John. Natural Law and Natural Rights. New York: Oxford University Press, 1980.

Galeotti, Anna E. Toleration as Recognition. New York: Cambridge University Press, 2002.

George, Robert P. Natural Law, Liberalism, and Morality: Contemporary Essays. New York: Oxford University

Press, 1996.

Gill, Emily R. Becoming Free: Autonomy and Diversity in the Liberal Polity. Lawrence: University Press of Kansas,

2001.

———. "Religious Organizations, Charitable Choice, and the Limits of the Freedom of Conscience." Perspectives

on Politics 2, no. 4 (2004): 741-55.

Gutmann, Amy. Democratic Education. Rev. ed. Princeton: Princeton University Press, 1999.

Habermas, Jurgen. An Awareness of What Is Missing: Faith and Reason in a Post-Secular Age. Translated by Ciaran

Cronin. Malden, MA: Polity, 2010.

Halbertal, Moshe. "Autonomy, Toleration, and Group Rights." In Toleration : An Elusive Virtue, edited by David

Heyd, 242. Princeton, NJ: Princeton University Press, 1996.

Hamburger, Joseph. John Stuart Mill on Liberty and Control. Princeton: Princeton University Press, 1999.

Hauerwas, Stanley. A Community of Character : Toward a Constructive Christian Social Ethic. Notre Dame, Ind. ::

University of Notre Dame Press,, c1981.

Heyd, David. "Introduction." In Toleration: An Elusive Virtue, edited by David Heyd, 242. Princeton, NJ: Princeton

University Press, 1996.

———. "Is Toleration a Political Virtue?" 2004.

———. "Is Toleration a Political Virtue?" In Toleration and Its Limits, edited by Melissa Williams, 171-94. New

York: New York University Press, 2008.

Horton, J. "Toleration as a Virtue." In Toleration: An Elusive Virtue, edited by David Heyd, 242. Princeton, NJ:

Princeton University Press, 1996.

Kalyvas, Stathis N. "Commitment Problems in Emerging Democracies: The Case of Religious Parties."

Comparative Politics